1. Dr Paul Mee – London School of Hygiene and Tropical Medicine
Does the use of HIV self-testing kits lead to unintended effects? Evidence from female sex workers in Malawi
Dr Paul Mee – London School of Hygiene and Tropical Medicine
HIV SELF-TESTING AFRICA
STAR Initiative

2. The authors declare that they have no competing interests.

3. What I will cover today Background to the study
Summary of research methods and analytical approach
Evidence for regret associated with HIVST use
Evidence for social harms following HIVST use
Conclusions and next steps

4. Background - HIV Self testing in Female Sex Workers
Worldwide Female Sex Workers (FSW) experience high levels of violence 1
RCTs show HIVST associated with an increased uptake and frequency of testing 2
HIV Self testing (HIVST) - a safe and acceptable alternative to Facility based testing for FSWs in Zambia, Uganda, Kenya & Zimbabwe 3,4,5,6
High readiness for HIVST in the General Population in Malawi – preferred to facility based testing 7
Concerns remain about social harms of HIVST amongst vulnerable groups
References
1 Deering K.N. et al . AMJPH (2014 ) 5 Thirumurthy H. et al.  Lancet HIV  (2016)
2 Johnson CC et al JIAS(2017) Napierela et al Bull WHO (in press)
3 Chanda, M.M. et al.  PLoS medicine (2017) 7 Choko A.T. et al PLoS medicine. (2015)
4 Ortblad, K. et al. PLoS medicine  (2017)

5. This study - Peer distribution of HIVST to FSW in Malawi
Research led by the team from the Malawi-Liverpool-Wellcome Trust Clinical Research Programme in Blantyre, Malawi (March – December 2017)
Part of a programme of studies into HIVST distribution in Malawi (Kumwenda M.K. et al. JIAS 2019)
Study goals
Investigate appropriate HIVST delivery models among FSW
Monitor for potential social harms
FSWs recruited as peer distributors – 5281 kits distributed , 2001 in Blantyre district
Prizes for most successful distributors
Participants could receive multiple kits over 3 months
Oral Fluid Test HIVST– OraQuick ADVANCE (with WHO pre-qualification approval)
Study protocol (

6. T= 0 – HIV Self test received
Data Collected
For each HIVST recipient
T= 0 – HIV Self test received
Baseline interview
Audio Computer Assisted Interviews (ACASI) interview
Collected data on socio-demographics,
sexual behaviour, testing history , social harms

7. T= 0 – HIV Self test received Three month longitudinal diary
Data Collected
For each HIVST recipient
T= 0 – HIV Self test received
Baseline interview
Audio Computer Assisted Interviews (ACASI) interview
Collected data on socio-demographics,
sexual behaviour, testing history , social harms
Three month longitudinal diary
Daily reports on sexual behavior & social harms
Weekly reports on HIV testing

8. T= 0 – HIV Self test received Three month longitudinal diary
Data Collected
For each HIVST recipient
T= 0 – HIV Self test received
Baseline interview
Audio Computer Assisted Interviews (ACASI) interview
Collected data on socio-demographics,
sexual behaviour, testing history , social harms
T= 3 months
Endline interview
Repeat ACASI interview
Changes in sexual behaviour, testing and
occurrences of social harms
Three month longitudinal diary
Daily reports on sexual behavior & social harms
Weekly reports on HIV testing

9. Analytical Questions from the endline data
What percentage of participants reported regret about HIVST use or relationship problems associated with HIVST use?
Did this differ by socio-demographic characteristics, test result or test initiator?

10. Characteristics of the study cohort
Variable
Category
N (%)
Age
16-25
151(57.0)
26-35
96(36.2)
>36
18(6.8)
Education level
Primary or less
170(64.2)
Secondary or higher
95(35.8)
Total
265
High rates of Intimate Partner Violence reported at baseline (48.4%)
Data on regrets and relationship problems associated with HIVST use available for 131/265 (49.4 %)
This group slightly older than those not reporting this data

11. Regret and relationship problems amongst those reporting HIVST use at endline
Percentage reporting regret about HIVST use was low and decreased over time since test
Variable
Category
Total
Immediate regret about HIVST use
Regret now about HIVST use
Relationship
problems caused by HIVST use %
p-value
Test initiator
Self 87
12.6
0.20
9.2
1.0
3.4
0.06
Other 44
22.7
9.1
13.6
HIVST result
Reactive 45
15.6
1.00
6.7
0.54
8.9
0.25
Unreactive 86
16.3
10.5
5.8
Age in years
16-25 66
18.2
0.40
16.7
< 0.01
0.51
26-35 53
11.3
3.8
>36 12 25
8.3
131
16.0 -
6.9

12. Immediate regret about HIVST problems caused by HIVST
ACASI results III – regret and relationship problems amongst those reporting HIVST use at endline
Evidence that relationship problems were greater if someone else initiated the HIVST
(mainly peer distributors)
Variable
Category
Total
Immediate regret about HIVST
Regret now about HIVST
Relationship
problems caused by HIVST %
p-value
Test initiator
Self 87
12.6
0.20
9.2
1.0
3.4
0.06
Other 44
22.7
9.1
13.6
HIVST result
Reactive 45
15.6
1.00
6.7
0.54
8.9
0.25
Unreactive 86
16.3
10.5
5.8
Age in years
16-25 66
18.2
0.40
16.7
< 0.01
0.51
26-35 53
11.3
3.8
>36 12 25
8.3
131
16.0 -
6.9

13. Immediate regret about HIVST problems caused by HIVST
ACASI results III – regret and relationship problems amongst those reporting HIVST use at endline
No evidence that regret or relationship problems were associated with HIVST result
Variable
Category
Total
Immediate regret about HIVST
Regret now about HIVST
Relationship
problems caused by HIVST %
p-value
Test initiator
Self 87
12.6
0.20
9.2
1.0
3.4
0.06
Other 44
22.7
9.1
13.6
HIVST result
Reactive 45
15.6
1.00
6.7
0.54
8.9
0.25
Unreactive 86
16.3
10.5
5.8
Age in years
16-25 66
18.2
0.40
16.7
< 0.01
0.51
26-35 53
11.3
3.8
>36 12 25
8.3
131
16.0 -
6.9

14. Longitudinal diary
Weekly reports of HIV self and facility based testing and coercion to test or disclose test result
Daily reports of type of sexual partners , condom use and social harms per partner

15. Longitudinal diary data

16. Longitudinal diary data
Analytical Questions i) What evidence is there that social harms are more likely in the week of a test and the following week compared to weeks where no tests occur and following week? ii) Is there evidence that this differs between self tests and facility based tests?

17. Results of GEE analysis
Including all individuals with baseline ACASI data (n=222)
Multivariate model using Generalised Estimating Equations (GEE) to account for correlated data within individuals and over time
Some evidence for increase in reports of verbal abuse in week of HIVST and following week
Type of social harm
Adjusted 1 odds ratio (95%CI) For occurrence of social harm in week of test and following week compared to weeks with no test and following week
Any test
Clinic Test only
Self Test only
Verbal
1.22 (0.90 – 1.66)
 1.32 (0.76 – 2.29)
 1.42 (0.98 – 2.07)
Physical
1.07 (0.76 – 1.49)
 0.84 (0.43 – 1.63)
 1.27 (0.84 – 1.91)
Sexual
1.02 (0.72 – 1.43)
 0.94 (0.49 – 1.79)
 1.16 (0.76 – 1.77)
Economic
0.95 (0.70 – 1.31)
 1.23 (0.71 – 2.15)
 0.91 (0.61 – 1.36)
1 Adjusted for total sex encounters per week, age and self identification as sex worker

18. Summary and Conclusions
Peer distribution strategy effective in reaching FSW with HIVST
Percentage of FSW reporting regret about taking HIVST was low and decreased over time
Regret not associated with test result
Further qualitative analysis probing meaning behind “regret”
Relationship problems more likely to be reported if test initiated by another person (mainly peer distributor)
Therefore important to consider alternative models to peer distributors for HIVST distribution among FSW
Some evidence in this small study population for increase in reports of verbal abuse following HIVST use for peer distribution model
Based on this and other studies – It is important to consider HIVST as an alternative where safe, acceptable and effective for reaching PLHIV in key at risk populations

19. Acknowledgements Study participants STAR project funders - Unitaid
Co-authors - Melissa Neuman1, Moses Kumwenda2, Mwiza Sambo2, Wezzie Lora 2 ,Pitchaya Indravudh1,2, Karin Hatzold3, Cheryl Johnson4, Liz Corbett1,2, Nicola Desmond2
1 London School of Hygiene and Tropical Medicine, UK 2 Malawi-Liverpool Wellcome Trust Clinical Research Programme, Blantyre, Malawi 3 Population Services International, Johannesburg, South Africa 4 Department of HIV/AIDS, WHO Geneva, Switzerland